Doctorado en Ingeniería Mecánica

URI permanente para esta colección

https://hdl.handle.net/10893/36606

Navegar

Examinando Doctorado en Ingeniería Mecánica por Título

Mostrando 1 - 1 de 1
  • Miniatura
    PublicaciónAcceso abierto
    Mechanical ventilatory parameters based on a mathematical model for diagnosis/treatment of older adults with ARDS
    (Universidad del Valle, 2025) Ruiz Hidalgo, Iván Dario; Jaramillo Pizarro, Guillermo Andrés; garcia melo, jose isidro; CIENCIA, EDUCACIÓN Y DIVERSIDAD
    This study presents a novel mathematical model to estimate mechanical ventilatory parameters in older adults diagnosed with Acute Respiratory Distress Syndrome (ARDS). The proposed model, called the Recruitment and Distention Elastance Analysis + Slice model (RDEA + Slice), builds on the traditional Single Compartment Model (SCM) by incorporating new strategies for improved physiological representation. Validation with real patient data showed that RDEA + Slice offers enhanced numerical accuracy compared to conventional methods, especially during spontaneous breathing and patient-ventilator asynchronies. To further explore asynchrony detection, the pressure signal was interpreted as a combination of a synchronous baseline and "noise" from asynchronous activity. A frequency-domain filtering approach was used to isolate these components, leading to the formulation of two metrics: the Asynchrony Event Percentage (AE%) and the Asynchrony Index (AI%). These metrics were applied to three datasets, allowing for the identification of threshold-based criteria to distinguish between synchronous and asynchronous breathing patterns. Additionally, image processing techniques were validated and used to extract ventilatory data from video recordings of ventilator screens. This approach provides a practical solution for data acquisition in low-resource settings, a common challenge in emerging economies. The extracted data showed high fidelity and low Mean Squared Error when compared with sensor-derived measurements. The results suggest that the proposed model, detection indices, and image processing methodology could support clinical decision-making by providing insights into patient-specific respiratory mechanics.